Quantification of urban structure on building block level utilizing multisensoral remote sensing data
Quantification of urban structure on building block level utilizing multisensoral remote sensing data
复制标题
利用多传感器遥感数据对城市结构进行街区级别的量化
作者：
M. Wurm;H. Taubenböck;S. Dech
城市环境的动态变化是对可持续发展的挑战。城市地区承诺财富、实现个人梦想和权力。因此，许多城市的特点是人口增长和物质发展。传统的可视化绘制和更新城市结构信息是一项非常费力和耗费成本的任务，对于大城市地区来说尤其如此。为此，我们开发了一种通过基于对象的图像分类来提取相关信息的工作流程。以这种方式，通过数字地面模型结合高度信息和非常高分辨率的光学卫星图像对多遥感遥感数据进行了分析，以检索具有相关土地利用/土地覆盖信息的详细3D城市模型。这些信息已汇总在建筑单元的层面上，以实物指标描述城市结构。已完成分类得出的指标与参考分类之间的比较，以显示单个指标与城市结构类型参考分类之间的相关性。这些指示器已被用于应用集群分析以将各个区块分组为相似的集群。
Dynamics of urban environments are a challenge to a sustainable development. Urban areas promise wealth, realization of individual dreams and power. Hence, many cities are characterized by a population growth as well as physical development. Traditional, visual mapping and updating of urban structure information of cities is a very laborious and cost-intensive task, especially for large urban areas. For this purpose, we developed a workflow for the extraction of the relevant information by means of object-based image classification. In this manner, multisensoral remote sensing data has been analyzed in terms of very high resolution optical satellite imagery together with height information by a digital surface model to retrieve a detailed 3D city model with the relevant land-use / land-cover information. This information has been aggregated on the level of the building block to describe the urban structure by physical indicators. A comparison between the indicators derived by the classification and a reference classification has been accomplished to show the correlation between the individual indicators and a reference classification of urban structure types. The indicators have been used to apply a cluster analysis to group the individual blocks into similar clusters.